Waste Management 62 (2017) 43–51

Contents lists available at ScienceDirect

Waste Management
journal homepage: www.elsevier.com/locate/wasman

Earthworms change the quantity and composition of dissolved organic

carbon and reduce greenhouse gas emissions during composting
Abebe Nigussie a,b,c,⇑, Sander Bruun b, Andreas de Neergaard b, Thomas W. Kuyper a
a
Department of Soil Quality, Wageningen University and Research, PO Box 47, 6700 AA Wageningen, The Netherlands
b
Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark
c
Department of Natural Resource Management, Jimma University, Ethiopia

a r t i c l e i n f o a b s t r a c t

Article history: Dissolved organic carbon (DOC) has recently been proposed as an indicator of compost stability. We
Received 25 October 2016 assessed the earthworms’ effect on DOC content and composition during composting, and linked compost
Revised 3 February 2017 stability to greenhouse gas emissions and feeding ratio. Earthworms reduced total DOC content, indicat-
Accepted 6 February 2017
ing larger stability of vermicompost than of thermophilic compost. The concentrations of humic acid and
Available online 20 February 2017
fulvic acid were reduced by earthworms, whereas there was no significant effect on hydrophobic neutrals
and hydrophilics. The humic acid fraction was depleted more quickly than the other compounds, indicat-
Keywords:
ing humic acid degradation during composting. The optimum feeding ratio decreased DOC content com-
Vermicomposting
Nitrous oxide
pared to the high feeding ratio. The lowest N2O emissions were also observed at the optimum feeding
Feeding ratio ratio. Our study confirmed the use of DOC content and composition as an indicator of compost stability
Eisenia fetida and suggested that feeding ratio should be considered when assessing the earthworms’ effect on stabil-
Dendrobaena veneta isation and greenhouse gas emissions.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction pathogens (Edwards, 2011), but the mechanisms of how earth-

Thermophilic composting and vermicomposting are two com-
posting techniques commonly used to convert biodegradable
waste into compost (Lazcano et al., 2008; Nigussie et al., 2016).
Thermophilic composting is a microbially-mediated,
high-temperature (>45 °C) process, while vermicomposting is a
mesophilic (<30 °C) process that involves earthworms and associ-
ated microorganisms in the decomposition and stabilisation of
organic materials (Munroe, 2007). The temperature during
vermicomposting should remain within the range of 15–30 °C, as
temperatures above 35 °C kill earthworms (Munroe, 2007).
Considerable decomposition while retaining higher nutrient
concentrations was observed during vermicomposting compared
with thermophilic composting (Nigussie et al., 2016; Lazcano
et al., 2008). In contrast, high N losses occur during thermophilic
composting because high temperatures (>45 °C) increase ammonia
volatilisation (Pagans et al., 2006). Temperatures above 45 °C were
considered essential for eradicating weeds and pathogens from
compost (Ryckeboer et al., 2003), however, vermicomposting has
also been shown to be effective at eradicating weeds and
⇑ Corresponding author at: Department of Plant and Environmental Sciences,
University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark.
E-mail addresses: nigatu@plen.ku.dk, abebe.nigatu@wur.nl (A. Nigussie).
worms kill weed seeds and pathogens is not known and the reports
are contradictory. Hence the combination of thermophilic com-
posting and vermicomposting has been proposed to produce com-
post of high agronomic value and pathogen-free (Lazcano et al.,
2008). Generally, the combination also enables the organic fertil-
izer to be produced at a faster rate than either of the individual
process (Lazcano et al., 2008). The first phase – thermophilic com-
posting – occurs only for a short period of time, mainly to eradicate
pathogens and eliminate toxic compounds, and the subsequent
vermicomposting (i.e. the second phase) is carried out to accelerate
the stabilisation process and improve the agronomic value of com-
post (Lazcano et al., 2008).
It is important that compost is sufficiently stable prior to soil
application because unstable compost reduces plant growth.
Unstable compost leads to oxygen depletion in the root zone,
osmotic stress and contains phytotoxic compounds (Wichuk and
McCartney, 2010). Compost is considered stable when the organic
matter decomposition rate is reduced to a low level with no heat
development. A number of indices are used to determine compost
stability (Bernal et al., 2009; Khan et al., 2014). Evolution of CO2 is
the most commonly used indicator (Bernal et al., 2009; Nigussie
et al., 2016), but this index is influenced by a number of factors
such as substrate quality. Lack of heat development is another sim-
ple method for evaluating compost stability (Boulter-Bitzer et al.

http://dx.doi.org/10.1016/j.wasman.2017.02.009
0956-053X/Ó 2017 Elsevier Ltd. All rights reserved.
44 A. Nigussie et al. / Waste Management 62 (2017) 43–51

2006), however it is also affected by aeration, pile size, moisture
content, degree of insulation and other parameters. Therefore the
use of one index to determine compost stability is potentially mis-
leading. Indices such as a C/N ratio <12 and a NH+4-N: NO
3 -N ratio
<0.16 are also recommended as a threshold level for indicating
compost maturity (Khan et al., 2014). Recently, studies showed
that the stability and maturity of the compost could be determined
by spectroscopy, structural characterization and thermogravimet-
ric analysis (Kumar et al., 2013). Numerous researchers have there-
fore suggested the combined use of multiple indices as indicator
for compost stability (Bernal et al., 2009; Khan et al., 2014).
Dissolved organic carbon (DOC) has recently been proposed as
an additional indicator of compost stability (Bernal et al., 2009;
Santos et al., 2016). A maximum threshold DOC value of 4 g kg
1
dry matter is used as an indicator of stable compost (Khan et al.,
2014). Not only the quantity but also the quality (i.e. chemical
composition) of DOC can be used to assess compost stability. A
batch fractionation procedure (Van Zomeren and Comans, 2007)
is currently used to separate DOC into four fractions, viz., humic
acid (HA), fulvic acid (FA), hydrophobic neutral (HON), all consid-
ered as hydrophobic compounds, and hydrophilic (Hi) compounds
(Straathof and Comans, 2015). A recent study has shown that the
proportions of these four fractions vary between composts, inde-
pendent of DOC concentration (Straathof and Comans, 2015). Hi
compounds declined during composting likely because they were
used as a substrate for microorganisms, and the hydrophobic com-
pounds (HA, FA and HON) fractionally increased in stable compost
(Straathof and Comans, 2015).
It is plausible that earthworms influence the DOC quantity and
quality (composition) of compost because they ingest the sub-
strates and thereby condition the microbial communities that
influence the decomposition process. Previous studies have found
higher stabilisation of compost as a result of vermicomposting
compared to thermophilic composting using indices such as CO2
evolution (Nigussie et al., 2016; Ngo et al., 2013) and biochemical
analysis (Lazcano et al., 2008; Ravindran and Mnkeni, 2016). How-
ever, little is known about the effect of earthworms on the quantity
and composition of DOC during vermicomposting.
Feeding ratio is defined as the ratio of substrate added over
earthworm biomass (Ndegwa et al., 2000). A high feeding ratio
decreases the conversion rate of fresh materials into vermicom-
post. Previous studies have shown that very high food supply
reduces the biomass and reproduction of earthworms (Luth et al.,
2011). Furthermore, Ndegwa et al. (2000) found that low feeding
ratio increases the mineralisation of nutrients (particularly nitro-
gen) compared with high feeding ratio. High feeding ratio increases
temperature and impedes air circulation in the pile (Luth et al.,
2011), both of which affect GHG emissions. For instance, if food
supply is too high (supra-optimal feeding ratio) per unit earth-
worm biomass, the temperature in the pile increases; high temper-
atures and anoxic patches not only result in increased earthworm
mortality, but in greater GHG emissions as well. Feeding ratio is
therefore an essential parameter that should be considered when
assessing the effect of earthworms on stabilisation and greenhouse
gas emissions. Recent reports have used substrate quality (Nigussie
et al., 2016; Wang et al., 2014) and earthworm density (Nigussie
et al., 2016) to evaluate the effect of earthworms on decomposition
and GHG emissions. In addition, feed type affects the conversion
rate of fresh materials into vermicompost (Edwards and Bohlen,
1996). However, the effect of feeding ratio on stabilisation pro-
cesses and GHG emissions during vermicomposting is not known.
The objectives of the present study were therefore (i) to evaluate
the effect of earthworms on DOC quantity and composition of
the compost, linking this effect to the initial substrate quality
and feeding ratio, and (ii) to assess the effect of feeding ratio on
GHG emissions from vermicomposting. We hypothesised that (i)
earthworms reduce the DOC content of compost compared to
non-earthworm composting, with the effect of earthworms being
greater at the optimal feeding ratio; (ii) earthworms reduce the
fractional contribution of Hi and hence increase the fractional con-
tribution made by HA, FA and HON compared to non-earthworm
composting, (iii) high feeding ratio increases GHG emissions from
vermicomposting compared to the optimal feeding ratio, and (iv)
high feeding ratio reduces compost stability, as assessed by CO2
flux, compared to the optimal feeding ratio.
2. Materials and methods
2.1. Experimental setup
Pre-decomposed garden waste was obtained from Unifarm, part
of Wageningen University and Research, and placed in plastic
boxes (30 cm width  40 cm length  25 cm height). Three sub-
strates with different composition that have undergone different
degrees of decomposition were used as composting materials.
The first substrate (substrate_1) was pre-composted for three
months, and the second substrate (substrate_2) was pre-
composted for nearly 1½ months. The third substrate (substrate_3)
was prepared from substrate_1, substrate_2 and cattle manure at a
ratio of 1:1:1 (weight basis). The cattle manure was obtained from
Unifarm, and added to the third substrate to increase nitrogen
availability in the mixture, whereas pre-decomposed materials
were used in this experiment to avoid the development of high
temperatures in the vermicompost bins. Hence there was no tem-
perature effect in our experiment unlike previous composting
experiments (Nigussie et al., 2016; Straathof and Comans, 2015).
Mixtures of adult individuals of two common composting
earthworm species, namely Eisenia fetida and Dendrobaena veneta
(approximate 2:1 ratio), were obtained from two earthworm
breeding companies, ‘De Polderworm’ and ‘Star Foods’, the Nether-
lands. The earthworms were added at a stocking density of
3 kg earthworms m
2. The substrates were added to the vermi-
composting bin in two doses: (i) 1.5 kg substrate kg earthworms
1
(recommended by Aira and Domínguez, 2008) – hereafter referred
to as optimal ratio (OR) – and (ii) 3 kg substrate kg earthworms
1 –
hereafter referred to as the high ratio (HR). Treatments without
earthworms were used as controls. The experiment had two factors
arranged in a 3  3 (earthworm treatments (OR, HR and control) 
substrate quality) complete randomised design with three repli-
cates. The experiment was conducted for 60 days, and the moisture
content in each container was adjusted approximately to 70–75%
by spraying of water on top.
2.2. DOC fractionation
DOC was extracted using ultra-pure water, as described by
Straathof and Comans (2015). Briefly, fresh compost was mixed
with ultra-pure milli-Q water at a 1:10 ratio (w/v), shaken for
one hour on a horizontal shaker and filtered through a 0.45 lm fil-
ter (WhatmanTM). Due to the heterogeneity of the compost samples,
each sample was replicated four times and the replicates were
finally pooled after the extracts had been filtered. A sub-sample
(5 ml) was then taken and analysed for DOC concentration using
San++ channel SFA (SKALAR, The Netherlands). The remaining sam-
ples were used for DOC fractionation.
The batch fractionation procedure (Van Zomeren and Comans,
2007) was used to separate the DOC fractions. Briefly, 40 ml of
the DOC sample was added in a 50 ml centrifuge tube, acidified
to pH 1.0 with 6 M HCl and allowed to stand overnight. This step
allowed the humic acid (HA) fraction to precipitate and form pel-
lets. The acidified solution was then centrifuged (20 min, 3500g)
 A. Nigussie et al. / Waste Management 62 (2017) 43–51 45

to separate the HA (i.e. the pellets) from the supernatant contain- NH+4 and NO
3 concentrations using segmented flow analysis
ing FA, HON and Hi (FaHiHON). About 25 ml of 0.1 M KOH (pH (SFA) (SKALAR analytical, the Netherlands).
12.0) was added to the pellets and shaken for 20 min to re-
suspend the HA fraction. The supernatant was transferred to
2.5. Statistical analyses
another 50 ml centrifuge tube. About 15 ml of the supernatant
was added to a 3 gm DAX-8 resin (Sigma–Aldrich), shaken for
A two-way analysis of variance (ANOVA) was performed to test
one hour and filtered through 0.45 lm (Whatman TM). This step
for significant effects of earthworm treatments, substrate quality
separated the hydrophilic compounds (Hi) from the supernatant.
and their interactions. A separate ANOVA was performed - exclud-
Finally, 15 ml 0.1 M KOH was added to the DAX-8 resin, shaken
ing the control treatment - to test for significant effects of two
for one hour, and filtered through 0.45 lm (WhatmanTM) to sepa-
feeding ratios on the vermicomposting process. The data were
rate the fulvic acid (FA) fractions. This step was repeated three or
checked for the assumptions of ANOVA prior to data analysis.
more times until the concentration in the samples was equal to
Levene and Shapiro-Wilk’s tests were used to test for homogeneity
the blank samples. The neutral (HoN) fractions were estimated
of variance and normality respectively. Data on the change in
from the DOC that was not dissolved under alkaline conditions
earthworm biomass did not fit with ANOVA assumptions and
(FaHiHON - (FA + Hi)). The concentrations of each fraction (i.e.
hence the data were log-transformed. Tukey test was used to com-
HA, Hi, FA and HoN) were measured using San++ channel SFA (SKA-
pare the means if the factors’ effect was significant at P < 0.05. Lin-
LAR, The Netherlands).
ear regression was performed between the CO2 emissions in the
last week as an independent variable and DOC quantity and DOC
2.3. Gas sampling composition as a dependent variable. All statistical analyses were
undertaken using SAS version 9.2.
The static chamber method was used to collect gas samples
(Chan et al., 2010). The gas samples were collected every two days
for the first week and then once a week until the end of the exper- 3. Results
iment, after 60 days. Gas samples were collected three times after
closing the chamber (0, 20 and 40 min). Gas samples were also col- 3.1. DOC quantity and composition
lected after 0, 20, 40, 60, 80 and 100 min every month in order to
check the linearity assumption. The gas samples were measured The effects of earthworms on DOC concentration are presented
using INNOVA 1412 photoacoustic field gas monitor (LumaSense in Fig. 1. There was a significant effect of earthworm treatment
Technologies, Ballerup, Denmark). The emissions rate in mg kg
1 (P < 0.001) and substrate quality (P < 0.001) on DOC concentration
initial dry matter day
1 was calculated as: as well as a significant interaction between them (P = 0.001). The





 presence of earthworms reduced DOC concentration by 7–28%,
DC V M P 273 depending on substrate quality and feeding ratio. The DOC concen-
Emission rate ¼     ð1Þ
Dt W Vs Po T tration per total C mineralised was calculated, and earthworms
decreased the total DOC concentration by 38–60% compared with
where DC is the change in concentration of gas (ppm) during time
the non-earthworm treatments. The effect of earthworms on
interval Dt in days, V is the headspace volume (litres), M is the
DOC was observed more in substrate_3. DOC contents were lower
molecular mass of the gas of interest (44, 16 and 44 g for CO2,
than 4 g kg
1 dry matter with substrate_1, around 4 g kg
1 dry
CH4 and N2O respectively), Vs is the volume occupied by 1 mol of
matter with substrate_2 and higher than the critical limit value
a gas at standard temperature and pressure (22.4 L), P is the atmo-
of 4 g kg
1 dry matter for substrate_3, indicating decreasing stabil-
spheric pressure (bar), Po is the standard pressure (1.013 bar), T is
ity of the final product. The optimal feeding ratio significantly
the temperature inside the chamber during the deployment time
(P < 0.01) decreased DOC concentration compared with the higher
in Kelvin, and W is the initial dry mass of the composting material
ratio.
(kg).
The percentage of DOC retrieved throughout the fractionation
The cumulative emissions were calculated using the trapezoid
procedure was 85–98%, which was comparable with the previous
integration rule (Ly et al., 2013).
study on compost (Straathof and Comans, 2015). Concentrations
ðtb
ta Þ:ðFta þ Ftb Þ of hydrophilic and hydrophobic (HA, FA, HON) fractions of DOC
AtðabÞ ¼ ð2Þ
2 varied between treatments and there was a significant earthworm
x substrate interaction (Table 1). FA was the dominant proportion
where At(ab) is the cumulative emission between the measurement
of DOC (>45%) while HON contributed the lowest proportion
days (between ta and tb), ta and tb are the dates of the two measure-
(<14%), irrespective of the treatments (Supplementary Fig. 1). Anal-
ments, and Fta and Ftb are the gas fluxes at the two measurement
ysis of variance showed that the concentrations of HA, FA, HON and
dates.
Hi were affected by substrate quality (P = 0.001, P = 0.001,
Therefore, the total cumulative emission was calculated as the
P < 0.001 and P < 0.001 respectively), but the presence of earth-
sum of cumulative emissions on each day using Eq. (3):
worms only had a significant effect on HA and FA (P = 0.001 and
X
Total cumulative emission ¼ AtðabÞ ð3Þ P = 0.001 respectively). Substrates that underwent the longest
pre-composting period (i.e. substrate_1) had the lowest concentra-
tions of HA, FA, HON and Hi. The relative proportions of each frac-
2.4. Chemical analyses tion to total DOC are presented in Supplementary Fig. 1. The
presence of earthworms decreased the relative proportion of HA
Compost samples were collected at the end of the experiment to total DOC (P = 0.02), but the relative proportions of FA, HON
for the analyses of pH, NO
+
3 and NH4. The samples were stored at and Hi were not significantly affected by the earthworms. Simi-
4 °C prior to laboratory analysis. pH was measured from a com- larly, feeding ratio did not affect the proportion of each fraction
post:water ratio of 1:10 (w/v), whereas NO
+
3 and NH4 concentra- to total DOC. Linear regression between the last week of CO2 evo-
tions were determined using 1 M KCl. The compost samples were lution and the concentration of DOC and its different fractions
mixed with 1 M KCl at a ratio of 1:100 compost:solution (w/v) (Fig. 2) showed a significantly positive relationship for all compo-
and shaken for one hour. The extracts were then analysed for nents (P < 0.001 in all cases). The stability of the various com-
46 A. Nigussie et al. / Waste Management 62 (2017) 43–51

6 Sources of variation P-value

DOC (g kg-1 DM)

Earthworms < 0.001
Substrate < 0.001
4 Earthworms*Substrate 0.01

0
CON OR HR CON OR HR CON OR HR
Substrate_1 Substrate_2 Substrate_3

Fig. 1. DOC concentration of the composts after 60 days of composting (mean + standard error of the mean; n = 3). CON = without earthworms, OR = optimal substrate-to-
earthworm ratio, HR = higher substrate-to-earthworm, DM = dry matter.

Table 1
Absolute concentration (mean + SEM) of the different DOC fractions after 60 days of composting.

Substrate DOC fractions (g kg
1 DM)

HA FA HON Hi
Substrate_1 CON 0.34 ± 0.07 1.13 ± 0.10 0.37 ± 0.03 0.38 ± 0.03
OR 0.20 ± 0.10 0.90 ± 0.10 0.23 ± 0.12 0.32 ± 0.01
HR 0.22 ± 0.02 0.94 ± 0.17 0.22 ± 0.18 0.31 ± 0.01
Substrate_2 CON 0.61 + 0.06 2.02 ± 0.09 0.41 ± 0.04 0.80 ± 0.07
OR 0.53 ± 0.09 1.75 ± 0.01 0.56 ± 0.04 0.67 ± 0.03
HR 0.52 ± 0.11 1.61 ± 0.01 0.60 ± 0.12 0.78 ± 0.04
Substrate_3 CON 1.15 ± 0.07 2.97 ± 0.23 0.62 ± 0.01 0.99 ± 0.10
OR 0.57 ± 0.10 2.79 ± 0.27 0.52 ± 0.06 0.95 ± 0.08
HR 0.82 ± 0.06 2.76 ± 0.11 0.62 ± 0.05 1.06 ± 0.04
ANOVA
Earthworm (E) <0.001 <0.001 0.42 0.49
Substrate (S) <0.001 <0.001 <0.001 <0.001
Earthworm * Substrate 0.007 0.18 0.01 0.03

CON = without earthworms, OR = optimal substrate-to-earthworm ratio, HR = high substrate -to-earthworm ratio, HA = humic acid, FA = fulvic acid, HON = hydrophobic
neutral, Hi = hydrophilic compounds, SEM = standard error of the mean, DM = dry matter.

pounds, as judged from the slope for the various compounds, was than the high feeding ratio (P < 0.001) when CO2 was accounted
different. With a 50% reduction of the final CO2 flux, HA was most for in the total GHG budget.
reduced (to 40%), followed by DOC, FA and Hi (51–53%), while HON
was most stable (reduced to 59%).
3.3. Chemical properties

The NO
3 concentration differed between the earthworm treat-

3.2. GHG emissions
The earthworm treatments and substrate quality influenced
GHG emissions during composting (Fig. 3). Total cumulative CO2
emissions differed between the earthworm treatments (P < 0.001)
and substrate quality (P < 0.001). There was also an interaction
between the two factors (P < 0.001). The presence of earthworms
increased CO2 production mainly from substrate_2 and sub-
strate_3. Similarly, the optimal feeding ratio increased CO2 emis-
sions from vermicomposting compared with the higher ratio
(P < 0.001). N2O emissions were affected by the earthworm treat-
ments (P = 0.001) and substrate quality (P < 0.001). The high feed-
ing ratio increased cumulative N2O emissions compared with the
optimal ratio (P < 0.001), but no difference was observed between
high feeding ratio and composting without worms. CH4 production
was very small in all treatments, and its contribution to the GHG
budget was negligible and there was no significant effect of the
treatments (data not shown). When CO2 was excluded from the
total GHG emissions, the optimum feeding ratio decreased total
GHG emissions compared to the high feeding ratio and non-
earthworm treatment. However, both earthworm treatments had
a higher total GHG budget than non-earthworm treatments
(P < 0.001) when CO2 was included in the total GHG budget. Simi-
larly, the optimum feeding ratio had higher total GHG emissions
ments (P < 0.001), substrate quality (P < 0.001) and there was also
a significant interaction (P = 0.02). The presence of earthworms
increased NO
3 concentration by up to 400% compared with the
non-earthworm treatments. NH+4 concentration was not, however,
affected by the earthworm treatments (P = 0.49), substrate quality
(P = 0.51) and there was no interaction effect (P = 0.71). The NH+4:
NO
3 ratio varied between the earthworm treatments (P < 0.001),
substrate quality (P = 0.02) and there was also a significant interac-
tion (P = 0.03). The earthworm treatments reduced the NH+4: NO
3
ratio by up to 50–80%. The pH was affected by the earthworm
treatments (P < 0.001) and substrate (P = 0.01). The presence of
earthworms reduced pH compared with non-earthworm treat-
ments, irrespective of the substrate quality. The optimal feeding
ratio resulted in lower pH compared to the high feeding ratio
(P < 0.01).
3.4. Earthworm biomass
The change in earthworm biomass was significantly affected by
substrate quality (P = 0.01), but not by feeding ratio (P = 0.03) or
the interaction (P = 0.24). The substrate that pre-decomposed for
nearly 1½ months resulted in the highest increase in earthworm
biomass (Fig. 5).
 A. Nigussie et al. / Waste Management 62 (2017) 43–51 47

1.4
y = 0.0063x - 0.0999 b
1.2
7 r² = 0.52

HA (mg kg DM)
y = 0.0365x + 0.0904 1
6 a

DOC (g kg-1 DM)

r² = 0.63

-1
5 0.8
4
0.6
3
0.4
2
1 0.2

0 0
0 50 100 150 200 0 50 100 150 200
CO2-C (mg kg-1 DM) CO2-C (mg kg-1 DM)

0.8
y = 0.0037x + 0.0728
0.7 d
3.5 r² = 0.62

HON (mg kg DM)

y = 0.0168x + 0.0991 0.6
3 r² = 0.56 c
FA (mg kg-1 DM)

0.5

-1
2.5
2 0.4
1.5 0.3
1 0.2
0.5 0.1
0 0
0 50 100 150 200 0 50 100 150 200
CO2-C (mg kg-1 DM) CO2-C (mg kg-1 DM)

1.4
y = 0.0066x + 0.0181
1.2 r² = 0.60 e
Hi (g kg -1 DM)

1
0.8
0.6
0.4
0.2
0
0 50 100 150 200
CO2-C (mg kg-1 DM)

Fig. 2. Linear regression between CO2-C emissions at the end of experiment and different dissolved organic carbon pools at the end of experiment: (a) total DOC, (B) humic
acid, (c) fulvic acid, (d) hydrophobic neutral, (e) hydrophilic compounds. DM = dry matter.

300 Sources of variations P-value

a
CO2-C (g kg-1 DM)

Earthworm < 0.001

200 Substrate < 0.001
Earthworm*substrate < 0.001

100

0
CON OR HR CON OR HR CON OR HR
Substrate_1 Substrate_2 Substrate_3

350
Sources of variations P-value
N2O-N (mg kg-1 DM)

300 b
Earthworm 0.001
250 Substrate < 0.001
Earthworm*substrate 0.96
200
150
100
50
0
CON OR HR CON OR HR CON OR HR
Substrate_1 Substrate_2 Substrate_3

Fig. 3. Total cumulative GHG emissions after 60 days of composting: (a) CO2 emissions, (b) N2O emissions. CON = without earthworms, OR = optimal earthworm-to-substrate
ratio, HR = higher earthworm-to-substrate ratio.
48 A. Nigussie et al. / Waste Management 62 (2017) 43–51
4. Discussion
4.1. DOC quantity and composition
The hypothesis that earthworms decreased the concentration of
DOC compared with the non-earthworm treatment was confirmed.
The result was consistent with the increase in C mineralisation
caused by the earthworms (Figs. 2a & 3a). In agreement with our
findings, Hanc and Dreslova (2016) and Lazcano et al. (2008) found
a lower DOC concentration after vermicomposting than after ther-
mophilic composting. A new vermicomposting parameter (i.e.
feeding ratio) was considered in the present study, and demon-
strated to have a significant effect on the DOC concentration
(Fig. 1). The optimum feeding ratio decreased the DOC concentra-
tion by 5–14% compared with the high feeding ratio. The effect of
feeding ratio on DOC was largest in the substrate that had under-
gone a shorter pre-composting period (substrate_3), and this is
mainly due to the high DOC content in the manure (17 g kg
1
dry matter) (data not shown). DOC contains easily degradable
compounds and its concentration declines consistently with com-
posting time. Similarly, it was found that composts that had under-
gone a longer pre-composting period (substrate_1) had a lower
DOC content. The significant correlation between DOC concentra-
tion and CO2 emissions from the last week of composting
(r2 = 0.63; P = 0.001) (Fig. 2a) confirmed active decomposition in
the substrates with a high DOC content. The DOC concentration
of compost influences microbial activities and C mineralisation
after soil application (Straathof et al., 2014) and hence it is used
as an indicator of compost stability. Composts with a low DOC con-
tent (less than 4 g kg
1 dry matter) are considered more stable
(Bernal et al., 2009), implying that the earthworms, especially at
optimum feeding ratios, stabilise compost more quickly than
non-earthworm treatments. The possible explanation for the
higher stabilisation caused by earthworms is that they enhance
the decomposition process through their interaction with microor-
ganisms. Substrate_3 was not fully stable because it had high total
DOC content (5–6 g kg
1 dry matter), and this is mainly due to the
presence of manure that underwent pre-decomposition for several
days. The DOC concentration observed in this study was compara-
ble with the earlier composting studies (Hanc and Dreslova, 2016),
which reported a DOC concentration in the range 3–6 g kg
1 dry
matter. However, Straathof and Comans (2015) reported a low
DOC content (<0.5–1.0 g kg
1) in some composts, although their
municipal-waste compost also contained high amounts of DOC
(5–7 g kg
1 dry matter). Their values cannot be directly compared
with our results, as in their composts organic materials were mixed
with soil prior to composting. After correcting for non-organic
(mineral) soil components, their compost contained between 2–
3 g kg
1 dry organic matter (compost with woody material or for-
est leaf litter) up to 19 g kg
1 dry matter (for municipal organic
waste). Current literature often reports the DOC concentration
per dry matter without considering the mineral fractions of com-
post materials (i.e. soils). Hence, we recommend that DOC be
expressed per weight of organic materials after discounting min-
eral content in order to assess compost stability. In our study, how-
ever, only organic materials were used and hence the DOC values
could be used directly to indicate compost stability.
The concentrations of different DOC compositions (FA, HA, HON
and Hi) was affected by the earthworm treatments, substrate qual-
ity and their interaction (Table 1). The presence of earthworms
decreased concentrations of FA and HA, whereas the concentra-
tions of all four components of DOC were lower in vermicompost
than in compost in the absence of worms. The low concentration
of the DOC fractions in the earthworm treatments coincided
with the decrease in the total DOC concentration (Fig. 1).
Therefore, the relative proportion of each fraction to total DOC
(Supplementary Fig. 1) could provide more reliable information
on the earthworms’ effect on DOC composition and the turnover
of the fractions during composting.
Straathof and Comans (2015) reported the DOC composition of
composts produced from different input materials and processing
conditions and noted significant variation in the relative contribu-
tion of HA, FA, HON and Hi in the various composts. Our study
showed only small effects of treatments and substrate on the frac-
tional contribution to total DOC. One major explanation for the dis-
crepancy is that these authors used a larger range of input material
(from wood and forest leaf litter to municipal waste) and there
were also large differences between composting practices (temper-
ature, time).
We hypothesised, in agreement with Straathof and Comans
(2015) that during the composting process, Hi compounds are
depleted more quickly than the three hydrophobic compounds
because the hydrophilic fraction consists of low-molecular-
weight sugars and amino acids, which are readily available C
sources for microorganisms. Straathof and Comans (2015) reported
a lower Hi proportion, but higher proportions of HA, FA and HON in
composts that had undergone a longer composting period and high
temperature (>70 °C) compared with composts produced at a
lower temperature with a composting period of <28 days (i.e.
unstable compost). They suggested that the fractional contribution
of Hi to DOC may be a better indicator for the contribution of com-
post on microbial activity. However, our data contradict this
hypothesis. In fact, the fractional contribution of HA declined with
increasing stability; and a ranking of the compounds with increas-
ing stability was HA > DOC  FA  Hi > HON. We suggest two
mechanisms to explain why HA turned out to be the least stable
component. First, decline of HA (and FA to a smaller extent) is
likely indicative for ligninolysis, whereas decline of Hi is indicative
for cellulolysis. While it has for a long time been taken for granted
that degradation of lignin proceeds at a slower rate than degrada-
tion of cellulose, recent data have actually indicated a faster
decomposition of lignin than cellulose (Klotzbücher et al., 2011).
Secondly, but equally important, changes in the various pools are
driven both by depletion of the compound during decomposition
but also by novel production during degradation and/or changes
in solubility of compounds. Van Zomeren and Comans (2007) put
their study in the framework of a novel perspective on humic sub-
stances as supramolecular associations of compounds with rela-
tively small mass, and observed that at declining HA
concentrations the chances increased that part of the HA fraction
actually showed up in the FA fraction. Further research is needed
to evaluate both potential mechanisms. Considering that the con-
centrations of HA and FA, but not those of HON and Hi, were lower
in the vermicompost than in the non-worm compost, independent
of the total amounts of DOC, we hypothesise a role of earthworms
and their associated microbes in the breakdown of aromatic poly-
mers. Earthworms also secrete mucus, and these compounds may
be rich in carbohydrates and proteins (Pan et al., 2010) that may
also preferentially end up in the Hi fraction. The HON fraction
was not affected by the presence of earthworms and/or feeding
ratio, and it is mainly because this fraction contains aliphatic com-
pounds, including quinones (Straathof and Comans, 2015), which
are resistant against microbial degradation. Next to the fractional
contribution of individual compounds, the ratio of hydrophilic to
hydrophobic substances (HA + FA + HON) may provide indications
of compost stability. Our compost showed ratios between 0.20
and 0.30, while the study of Straathof and Comans (2015) indicated
ratios of 0.66–0.79 for unstable composts.
To our knowledge, no studies have been conducted to assess the
effect of earthworms on DOC composition. Further studies are rec-
ommended to advance understanding of the effect of earthworms
on the DOC pools using different parameters such as substrate
 A. Nigussie et al. / Waste Management 62 (2017) 43–51
quality (i.e. fresh organic material), earthworm species and earth-
worm density.
4.2. Greenhouse gas emissions
As hypothesised, earthworms increased CO2 emissions com-
pared with non-earthworm treatments, and the results were com-
parable with earlier studies (Chan et al., 2010; Nigussie et al.,
2016). Similarly, the CO2 emissions from vermicomposting were
higher at optimum feeding ratio than at the high ratio, as hypoth-
esised. Higher cumulative CO2 emissions indicated a greater stabil-
ity of compost and confirmed that the earthworms resulted in
compost that was at a more advanced stage of stabilisation. The
CO2 results were in agreement with the observation on DOC quan-
tity (Fig. 1).
The hypothesis that the high feeding ratio increases N2O emis-
sions from vermicomposting was confirmed. As compared to the
non-earthworm treatment, earthworms decreased N2O emissions
by 23–48% at the optimum feeding ratio. At the higher ratio, how-
ever, earthworms had no significant effect on N2O emissions com-
pared with the non-earthworm treatments. The low N2O
production in the earthworm treatments could be explained by:
(i) continuous turning of substrates by earthworms which subse-
quently increases aeration (Chan et al., 2010; Nigussie et al.,
2016) and (ii) high substrate stability after it has passed through
the earthworms’ gut (Luth et al., 2011). Hence, the low DOC con-
centration as evidence for compost stability (Fig. 1) could explain
the low N2O emissions in the earthworm treatments. At the higher
feeding ratio, however, the presence of anaerobic patches and the
lower degree of stabilisation (Fig. 1) reduced the mitigation effect
of earthworms on N2O emissions.
Reports on the effect of earthworms on N2O emissions are con-
tradictory. For instance, Hobson et al. (2005) and Lubbers et al.
(2012) reported earthworm-induced N2O emissions. In contrast,
Chan et al. (2010), Nigussie et al. (2016) and Wang et al. (2014)
found that earthworms decreased N2O emissions during compost-
ing. The contrasting results can be explained by differences in
earthworm species (i.e. difference in their feeding and burrowing
behaviours) (Lubbers et al., 2013), substrate quality (i.e. carbon
quality and nitrogen content) (Luth et al., 2011; Nigussie et al.,
2016), temperature (Nigussie et al., 2016) and the scale of the
experiment (Chan et al., 2010). In our study the decreasing effect
of earthworms on N2O emissions occurred at optimum feeding
ratio, while at a high feeding ratio earthworms did not decrease
N2O emission, implying that feeding ratio is an important parame-
ter for consideration when assessing the earthworms’ effect on
GHG emissions during composting. Denitrification in the earth-
worm gut was the main process contributing to N2O emissions in
the case of anecic earthworms (Lubbers et al., 2013). Composting
materials mostly contain high levels of nitrogen, hence the contri-
bution of denitrification occurring in the earthworm gut to total
N2O emissions was smaller than the contribution from denitrifica-
tion in the environment around the worms. Furthermore, the feed-
ing and burrowing behaviours of earthworms mostly used in soil
experiments (i.e. anecic earthworms) are different from compost
worms.
Two scenarios were used to assess the effect of earthworm
treatments on total GHG emissions. The first scenario excluded
CO2 because higher CO2 emissions indicate a greater stability of
the material. Including CO2 in the GHG balance would therefore
privilege composts that are not stabilised over the course of the
experiment, and therefore provide a biased assessment how earth-
worms affect the GHG balance (Chowdhury et al., 2014). Under this
scenario, the effect of earthworm treatments on the total GHG
emissions was similar to their effect on N2O emissions since the
contribution of CH4 was negligible. However, when CO2 was
49
included in the total GHG budget, the earthworm treatments
increased total GHG emissions compared with the non-
earthworm treatments. This variation was explained by the high
CO2 emissions in the earthworm treatments (Fig. 3a), and the large
contribution of CO2 to the total GHG budget in all treatments
(>80%). These results are comparable with previous studies on
composting (Andersen et al., 2010; Nigussie et al., 2016).
4.3. Properties of compost
Irrespective of substrate quality, the presence of earthworms
and optimum feeding ratio decreased pH, but increased the NO
3
concentration. The high NO
3 concentration in the earthworm
treatments and optimum feeding ratio was explained by the min-
eralisation of N from the organic materials and/or reduced N2
losses through denitrification (Fig. 3b). The NH+4 concentration
decreases during compositing due to ammonia volatilisation and
nitrification processes, while the NO
3 concentration increases
towards the end of the composting period. Hence, the NH+4: NO
3
ratio has been used to assess the maturity of compost (Bernal
et al., 2009), and the threshold value of <0.16 is an indicator of
mature compost, implying that all the composts were mature. Still
the earthworm treatments produced more mature compost than
the non-earthworm treatments. We used the total DOC content
and NH+4: NO
3 ratio to estimate compost stability, and both indices
yield the same result except for substrate_3. The presence of man-
ure increased the NO
3 concentration in substrate_3, consequently
reduced the NH+4: NO
+

3 ratio, suggesting that NH4: NO3 ratio
depends on the substrate quality and hence it may not be a reliable
index compared with DOC. The decrease in pH values in the earth-
worm treatments at optimum feeding ratio could be due high nitri-
fication in these treatments (Fig. 4a) and the production of organic
acids as a result of greater decomposition (Figs. 1 & 3a) (Lazcano
et al., 2008).
The data presented here confirmed a significant effect of feeding
ratio on the stabilisation of organic material (Fig. 1), GHG emis-
sions (Fig. 3) and properties of the end product (Fig. 4), however
(applied) researchers on vermicomposting have paid little atten-
tion to feeding ratio. Feeding ratio influences earthworm growth
(Fig. 5: Luth et al., 2011), aeration and temperature, subsequently
affecting decomposition, mineralisation, nutrient losses and GHG
emissions. Hence, it is suggested that this ratio be considered an
equally important parameter when evaluating the effectiveness
of earthworms in the stabilisation of organic materials to parame-
ters such as earthworm species, and substrate quality.
5. Conclusions
Earthworms accelerated the stabilisation of organic materials
compared with the non-earthworm treatments, as confirmed by
CO2 production, DOC concentration, DOC composition and NH+4:
NO
3 ratio. Earthworms decreased the total DOC concentration,
but they did not affect the relative composition of hydrophilic, ful-
vic acid and hydrophobic neutral fractions. The relative contribu-
tion by humic acid decreased by 15–45% in the earthworm
treatments, implying that the humic acid fraction was less recalci-
trant than commonly assumed, and was likely used as a substrate
by microorganisms. The hydrophilic-to-hydrophobic ratio was
consistent between the different stable composts, despite differ-
ences between input materials and processing conditions, and this
ratio could be used as an additional criterion to assess compost sta-
bility. This ratio is much higher unstable composts, implying that it
can be used as an indicator of compost stability. Fractionation of
DOC is therefore important for understanding the stabilisation of
organic waste. A higher (supra-optimal) feeding ratio reduced the
50 A. Nigussie et al. / Waste Management 62 (2017) 43–51

2000 Sources of variations P-value

a

NO3- (mg kg-1 DM)

1600 Earthworms < 0.001
Substrate < 0.001
Earthworm*substrate 0.02
1200

800

400

0
CON LR HR CON LR HR CON LR HR
Substrate_1 Substrate_2 Substrate_3

Sources of variations P-value

8 b
Earthworms 0.49
Substrate 0.51
NH4+ (mg kg-1 DM)

Earthworms*substrate 0.71
6

0
CON LR HR CON LR HR CON LR HR

8 Sources of variations P-value

c
Earthworms < 0.001
Substrate 0.01
Earthworms*substrate 0.3
7
pH

5
CON LR HR CON LR HR CON LR HR
Substrate_1 Substrate_2 Substrate_3

Fig. 4. Chemical properties of the composts after 60 days of composting (mean ± standard error of the mean; n = 3). (a) NO
+
3 (b) NH4(c) pH. CON = without earthworms,
OR = optimal substrate-to-earthworm ratio, HR = high substrate-to-earthworm ratio.

160
Change in earthworm biomass (%)

Sources of variation P-value

Earthworms 0.03
120 Substrate 0.01
Earthworms*substrate 0.24

80

40

0
OR HR OR HR OR HR
Substrate_1 Substrate_2 Substrate_3

Fig. 5. Change in earthworm biomass after 60 days of vermicomposting. LR = optimal substrate-to-earthworm ratio, HR = high substrate-to-earthworm ratio.

stabilisation process and increased N2O emissions by 30–50% com- Acknowledgements

pared with the optimum feeding ratio. Hence feeding ratio should
be considered as an important parameter when assessing the This study was conducted as part of a PhD thesis project sup-
earthworms’ effect on the stabilisation of organic materials and ported by the Agricultural Transformation by Innovation (AgTraIn)
GHG emissions during composting. Erasmus Mundus Joint Doctorate Programme, funded by the EACEA
 A. Nigussie et al. / Waste Management 62 (2017) 43–51
(Education, Audio-visual and Culture Executive Agency) of the
European Commission. The authors would like to thank Mr Jaap
Nelemans for laboratory assistance, Mr. Lou Derks from De Polder-
worm for providing the earthworms and technical advice, prof. Rob
Comans for helpful discussion on DOC fractionation and the inter-
pretation of the various fractions, and three anonymous reviewers
for constructive criticism on an earlier version of the manuscript.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.wasman.2017.02.
009.
References
Aira, M., Domínguez, J., 2008. Optimizing vermicomposting of animal wastes:
effects of manure application on C loss and microbial stabilisation. J. Environ.
Manage. 88, 1525–1529.
Andersen, J.K., Boldrin, A., Samuelsson, J., Christensen, T.H., Scheutz, C., 2010.
Quantification of greenhouse gas emissions from windrow composting of
garden waste. J. Environ. Qual. 39 (2), 713–724.
Bernal, M.P., Alburquerque, J.A., Moral, R., 2009. Composting of animal manures and
chemical criteria for compost maturity assessment. A review. Bioresour.
Technol. 100, 5444–5453.
Boulter-Bitzer, J.I., Trevors, J.T., Boland, G.J., 2006. A polyphasic approach for
assessing maturity and stability in compost intended for suppression of plant
pathogens. Appl. Soil Ecol. 34, 65–81.
Chan, Y.C., Sinha, R.K., Wang, W., 2010. Emission of greenhouse gases from home
aerobic composting, anaerobic digestion and vermicomposting of household
wastes in Brisbane, Australia. Waste Manage. Res. 29, 540–548.
Chowdhury, M.A., de Neergaard, A., Jensen, L.S., 2014. Potential of aeration flow rate
and bio-char addition to reduce greenhouse gas and ammonia emissions during
manure composting. Chemosphere 97, 16–25.
Edwards, C.A., 2011. Human pathogen reduction during vermicomposting. In:
Edwards, C.A., Arancon, N.Q., Sherman, R. (Eds.), Vermiculture Technology:
Earthworms, Organic Wastes and Environmental Management. CRC Press, Boca
Raton, pp. 249–261.
Edwards, C.A., Bohlen, P.J., 1996. Biology and Ecology of Earthworms. third ed.
Chapman and Hall, London, pp. 35–55.
Hanc, A., Dreslova, M., 2016. Effect of composting and vermicomposting on
properties of particle size fractions. Bioresour. Technol. 217, 186–189.
Hobson, A.M., Frederickson, J., Dise, N.B., 2005. CH4 and N2O from mechanically
turned windrow and vermicomposting systems following in-vessel pre-
treatment. Waste Manage. 25, 345–352.
Khan, N., Clark, I., Sánchez-Monedero, M.A., Shea, S., Meier, S., Bolan, N., 2014.
Maturity indices in co-composting of chicken manure and sawdust with
biochar. Bioresour. Technol. 168, 245–251.
Klotzbücher, T., Kaiser, K., Guggenberger, G., Gatzek, C., Kalbitz, K., 2011. A new
conceptual model for the fate of lignin in decomposing plant litter. Ecology 92
(5), 1052–1062.
51
Kumar, D.S., Kumar, P.S., Rajendran, N.M., Anbuganapathi, G., 2013. Compost
maturity assessment using physicochemical, solid-state spectroscopy, and
plant bioassay analysis. J. Agric. Food Chem. 61, 11326–11331.
Lazcano, C., Gómez-Brandón, M., Domínguez, J., 2008. Comparison of the
effectiveness of composting and vermicomposting for the biological
stabilization of cattle manure. Chemosphere 72, 1013–1019.
Lubbers, I.M., van Groenigen, K.J., Fonte, S.J., Six, J., Brussaard, L., van Groenigen, J.W.,
2013. Greenhouse gas emissions from soils increased by earthworms. Nat Clim.
Change 3, 187–194.
Luth, R.P., Germain, P., Lecomte, M., Landrain, B., Li, Y., Cluzeau, D., 2011. Earthworm
effects on gaseous emissions during vermifiltration of pig fresh slurry.
Bioresour. Technol. 102, 3679–3686.
Ly, P., Jensen, L.S., Bruun, T.B., de Neergaard, A., 2013. Methane (CH4) and nitrous
oxide (N2O) emissions from the system of rice intensification (SRI) under a rain-
fed lowland rice ecosystem in Cambodia. Nutr. Cycl. Agroecosyst. 97, 13–27.
Munroe, G., 2007. Manual of On-Farm Vermicomposting and Vermiculture, Organic
Agriculture Centre of Canada Pp 56.
Ndegwa, P.M., Thompson, S.A., Das, K.C., 2000. Effects of stocking density and
feeding rate on vermicomposting of biosolids. Bioresour. Technol. 71, 5–12.
Ngo, P.T., Rumpel, C., Ngo, Q.A., Alexis, M., Vargas, G.V., Gil, M., Dang, D.K., Jouquet,
P., 2013. Biological and chemical reactivity and phosphorus forms of buffalo
manure compost, vermicompost and their mixture with biochar. Bioresour.
Technol. 148, 401–407.
Nigussie, A., Kuyper, T.W., Bruun, S., de Neergaard, A., 2016. Vermicomposting as a
technology for reducing nitrogen losses and greenhouse gas emissions from
small-scale composting. J. Clean. Prod. 136, 429–439.
Pagans, E., Barrena, R., Font, X., Sánchez, A., 2006. Ammonia emissions from the
composting of different organic wastes, dependency on process temperature.
Chemosphere 62, 1534–1542.
Pan, X., Song, W., Zhang, D., 2010. Earthworms (Eisenia foetida, Savigny) mucus as
complexing ligand for imidacloprid. Biol. Fertil. Soils 46, 845–850.
Ravindran, B., Mnkeni, P.N.S., 2016. Identification and fate of antibiotic residue
degradation during composting and vermicomposting of chicken manure. Int. J.
Environ. Sci. Technol. http://dx.doi.org/10.1007/s13762-016-1131-.
Ryckeboer, J., Mergaert, J., Vaes, K., Klammer, S., de Clercq, D., Coosemans, J., Insam,
H., Swings, J., 2003. A survey of bacteria and fungi occurring during composting
and self-heating processes. Ann. Microbiol. 53, 349–410.
Santos, A., Bustamante, M.A., Tortosa, G., Moral, R., Bernal, M.P., 2016. Gaseous
emissions and process development during composting of pig slurry: the
influence of the proportion of cotton gin waste. J. Clean. Prod. 112, 81–90.
Straathof, A.L., Chincarini, R., Comans, R.N.J., Hoffland, E., 2014. Dynamics of soil
dissolved organic carbon pools reveal both hydrophobic and hydrophilic
compounds sustain microbial respiration. Soil Biol. Biochem. 79, 109–116.
Straathof, A.L., Comans, R.N.J., 2015. Input materials and processing conditions
control compost dissolved organic carbon quality. Bioresour. Technol. 179, 619–
623.
Van Zomeren, A., Comans, R.N.J., 2007. Measurement of humic and fulvic acid
concentrations and dissolution properties by a rapid batch procedure. Environ.
Sci. Technol. 41, 6755–6761.
Wang, J., Hu, Z., Xu, X., Jiang, X., Zheng, B., Liu, X., Pan, X., Kardol, P., 2014. Emissions
of ammonia and greenhouse gases during combined pre-composting and
vermicomposting of duck manure. Waste Manage. 34, 1546–1552.
Wichuk, K., McCartney, D., 2010. Compost stability and maturity evaluation - a
literature review. Canad. J. Civil Eng. 37 (11), 1505–1523.